INTRODUCTION
Trauma is the most common cause of death among children and infants, of which head in- juries account for some 60%. The mortality and morbidity rates concerning primary lesions and posttraumatic sequelae in patients with head in- juries have been considerably reduced by the advent of computed tomography (CT), which is still the examination technique of choice in the acute phase, thanks to the rapidity with which it can be performed, the ready availability of the imaging equipment and the absence of con- traindications (6, 25, 29). The drawback of this technique is the difficulty encountered in de- tecting smaller lesions, which are often located at the grey-white matter junction or in the vicin- ity of bone (e.g., temporal and frontal lobe poles, posterior fossa). In certain cases, the pa- tient’s clinical condition can be quite in contrast with the information yielded on their respective CT examinations (4, 7).
Due to its greater sensitivity in detecting these types of lesions, magnetic resonance im- aging (MRI) can be used as a complement to or even a substitute for CT in some instances (5, 10, 14, 16, 18, 19, 24, 26). This said, perform- ing an MR examination in the acute phase of trauma can prove somewhat difficult and en- tails a number of risks. These drawbacks,
which will be discussed briefly below, make MR a technique of secondary importance in the overall imaging evaluation of head injuries (12).
DRAWBACKS Intrinsic limits
The intrinsic limits of the MR technique con- sist in the absolute contraindication of examining subjects having ferromagnetic foreign bodies or electronic device implants (e.g., cardiac pace- makers). This problem becomes important in polytrauma patients, who may have acquired metal splinters from the traumatic event itself, and even more so in subjects with disorders of consciousness in whom it is not possible to re- construct an accurate medical history with regard to previous surgical or prosthetic implant proce- dures (22, 9). It is therefore necessary if possible to determine the compatibility of any metallic foreign materials that are known or suspected to be present before exposure of the patient to the strong magnetic fields inherent in MRI.
Semeiological limits
The semeiological limits of MRI during the acute phase of cranial trauma include its lower
2.3
MRI IN HEAD INJURIES
M. Gallucci, G. Cerone, M. Caulo, A. Splendiani, R. De Amicis, C. Masciocchi
sensitivity as compared to CT in identifying bony fractures of the skull, especially small ones, and acute intracranial bleeds (1, 30).
Whereas in CT the densitometric value of the blood is due primarily to the quantity of the haemoglobin protein component, in MRI the intensity of the signal depends in large part on
the magnetic properties of the iron contained in blood, or rather, on the electronic configura- tion that the hemoglobin iron assumes during evolution of the haemoglobin breakdown process (2). Therefore, although CT is clearly able to detect a hyperacute phase haemorrhage as an obvious area of hyperdensity due to the
Fig. 2.26 - Hyperacute mesencephalic haemorrhage. CT (a) shows the haemorrhage as a hyperdense area in comparison with the sur- rounding brain tissue. The MRI examination conducted in the acute phase shows the haemorrhage to be isointense on T1-weighted sequences (b) and hyperintense on T2-dependent, FLAIR and turbo spin echo (TSE) (c, d) sequences because blood cannot be dis- tinguished from oedema in this phase on the basis of MRI (i.e., oxyhaemoglobin).
a
b
c
d
high concentration of haemoglobin, MRI in this same phase has characteristically poor sen- sitivity, as the blood (oxyhaemoglobin) has not yet undergone metabolic transformation to a paramagnetic species (Fig. 2.26).
It is only in the acute period, after approxi- mately 12 hours after the event responsible for the bleed, that the first of a series of haemoglo- bin breakdown products starts to form in amounts that will alter the MR signal. This new species, deoxyhaemoglobin, causes a shorten- ing of T2 time, and, to a greater extent T2*, thus creating local inhomogeneities in the mag- netic field generated by the MR unit. On the other hand, red blood cell lysis and the block- age of the respiratory chain in haemorrhages cause an increase in the quantity of free water and therefore an increase in T2 and proton density signal, with consequent balancing of the T2 shortening, effect and a possible con- cealment of the hypointensity linked to the presence of intracellular deoxyhaemoglobin.
As the magnetic susceptibility effect and there- fore T2* shortening is directly proportionate to the square of the intensity of the static magnet- ic field, by using high field MR appliances and sequences particularly sensitive to T2* (e.g., gradient echo sequences, echo-planar sampling techniques) it is possible to resolve the problem and achieve dominant T2 shortening effects on imaging (i.e., focal reduction of the MR signal within subacute haemorrhages).
Another limitation posed by MRI is the fact that it requires relatively longer examination times than does CT. Time is critical in the diag- nostic management of critical cranial trauma patients. In addition, these patients frequently are unable to consciously cooperate for the long examination times inherent in MRI, there- by often resulting in motion artefacts. Recent progress in technology has largely made it pos- sible to overcome such limits with the use of high field equipment having ultrafast acquisi- tion sequences, that allows to obtain single slice images in less than one second (9, 20) (Fig.
2.27b).
However, it should be remembered that in acute head injury patients the fundamental question to be answered is whether emergency surgery is required. In almost all cases this question is answered by CT, which is efficient in depicting significant haematomas or frac- tures of surgical interest. MRI is doubtlessly more sensitive in identifying subtle haemor-
Fig. 2.26 (cont.) - An MRI examination conducted three days later shows an area of hyperintense signal on T1-weighted se- quences (e) and hypointense signal on T2-weighted sequences as a result of haemoglobin breakdown products (i.e., deoxy- haemoglobin). [a), b), c) axial CT; d), f) axial T2-weighted MRI; e) axial T1-weighted MRI].
e
f
rhagic collections and small intraaxial lesions related to the traumatic incident, but such find- ings have no relevance with regard to emer- gency treatment (17, 23, 27). In any case, these minor imaging findings can be detected using MRI once the immediate life-threatening prob- lems have been addressed and the patient has stabilized.
SEMEIOTICS
Primary intraaxial traumatic lesions
Contusions and laceration-contusions repre- sent approximately 45% of all head injuries.
They occur as a consequence of the impact of brain tissue against the inner table of the skull or dural insertion of the cranium. This trauma tends to cause related lesions in the poles of the frontal and temporal lobes and within the brainstem (22, 25). MRI is generally more sen- sitive in detecting these types of lesions than is CT due to the proximity of the traumatic alter- ation to the bony structures (i.e., CT beam hardening effects obscure this pathology) (15).
A distinction must be made between simple non-haemorrhagic contusions and haemor- rhagic or laceration-contusions.
Simple (non-haemorrhagic) contusions are ar- eas of tissue characterized by the presence of oedema in the acute phase and by necrotic-en- cephalomalacic evolution in the chronic phase.
MR is estimated to be up to 50% more sensi- tive than is CT in identifying simple contusions, which appear as hyperintense on T2-weighted sequences, typically affecting the surface layers of the cortical gyri. The sensitivity of MR is fur- ther increased by the use of FLAIR sequences;
T1-weighted sequences by comparison are of little benefit in the acute phase in simple con- tusions (Fig. 2.28).
Laceration-contusions (haemorrhagic contu- sions) differ from simple contusions by the pres- ence of a haemorrhagic zone within the contu- sion. In this type of lesion, CT can have greater specificity than MR, which is often not able to differentiate between haemorrhagic and non- haemorrhagic lesions (Figs. 2.26, 2.29, 2.30). As
Fig. 2.27 - Multiple contusive and lacerative haemorrhagic cere- bral foci in the acute phase. The T2-weighted sequence (a) shows multiple areas of signal hyperintensity within which haemor- rhagic components cannot be distinguished. The long scan time often results in motion artefacts. The echo-planar sequence (b), which is more sensitive to T2*-weighted tissue, enables the iden- tification of areas of hypointense signal caused by the presence of subacute haemorrhage (deoxyhaemoglobin). Motion artefacts can also be reduced by obtaining the acquisition in approxi- mately 500 msec. [a) T2-weighted MRI; T2*-weighted MRI].
a
b
mentioned previously, the use of high field ap- pliances with sequences sensitive to T2* in- crease MR’s specificity, thereby enabling the identification of haemoglobin breakdown prod- ucts within the haemorrhagic focus (Fig. 2.27).
Axonal injury is caused by axonal shear le- sions that occur with linear or torsional inertial forces, which are common in instances of high speed trauma. They have preferential sites in specific areas of the brain that are particularly
Fig. 2.28 - Cerebral contusions. T2 fast spin echo (FSE) and FLAIR sequences demonstrate that the latter prove more sensitive in identifying contusive foci, even of very small dimensions. The superior identification of the subdural haematoma with FLAIR se- quences is also noted. [a), c), d) T2-weighted MRI; b) FLAIR MRI].
a
b
c
d
sensitive to these kinds of force in trauma events, including: the white grey-white matter interface, the border between the superficial
mantle and the deep grey matter structures (junctional lesions), the midline commissures (especially the corpus callosum: commissural
Fig. 2.29 - Hyperacute cerebral haemorrhage. CT and T2- weighted MRI images show that in the hyperacute phase only CT is capable of showing the haemorrhagic component of the right temporal-insular shear-contusive focus. [a) axial CT; b) axial T2-weighted MRI].
Fig. 2.30 - Contusion-shearing injury. T2-weighted MRI se- quences highlight multiple grey-white matter junction axonal (a) callosal commissure lesions (b, c), lesions in the basal gan- glia (c, d) and brainstem lesions (e). However CT alone is able to demonstrate the presence of any haemorrhagic component (f). [a)-e) axial T2-weighted MRI; f) axial CT].
a
b
a
b
lesions) and the brainstem (7, 13). These le- sions are usually ovoid in shape with their lon- gitudinal axis directed parallel to the direction of nerve fibre bundles, and having dimensions
ranging from a few millimetres to 2 cm; these lesions are usually not haemorrhagic.
Junctional lesions are believed to be the most frequently observed type of traumatic cerebral
Fig. 2.30 (cont.).
e
f c
d
injury and are usually located in the poles of the frontal and temporal lobes. Neuropathological- ly, the lesion is characterized by the transection of the nerve cell’s axon, with subsequent Wal- lerian degeneration of the distal axon process and associated swelling of the nerve cell body.
Junctional lesions only affect the grey-white matter interface in the basal ganglia in 4.5% of cases and usually have a focus in the thalamus;
in these cases at least 50% of lesions are haem- orrhagic, in part because of the rich vascular- ization of such tissue. The explanation for this preferential injury site lies in the difference be- tween the relative kinetic energy of white mat- ter versus grey matter, partly due to their dif- ferent specific weights. Therefore, in the case of abrupt acceleration-deceleration, the two sub- components come to a stop at different times, thus resulting in linear tears at their interface.
MR is up to three times more sensitive than is CT in identifying such junctional lesions, typi- cally demonstrating focal hyperintense areas on T2-weighted sequences having an ovoid shape and located at the border area between white and grey matter (Fig. 2.30). The sensitivity of MR can be further enhanced by the use of FLAIR sequences, which typically make it pos- sible to visualize even the smallest lesions (Fig.
2.31). Once again, the use of high field magnets and T2*-dependent sequences makes it possi- ble to detect haemorrhagic components when present (Fig. 2.32). However, it must be re- membered that although the identification of this type of lesion is not important for reasons of emergency treatment, it can be a significant factor in later prognostic and legal analyses.
Commissural lesions affecting the axonal fi- bres of the corpus callosum typically occur in cases where the brain has been subjected to greater degrees of trauma, often of a torsional nature. This torsion injury results from differ- ential forces applied to the corpus callosum when a different rotational kinetic energy ob- tains within one cerebral hemisphere as com- pared to the contralateral one during accelera- tion-deceleration trauma. The falx cerebri can act as a contributory factor in the traumatic ef- fects, causing mechanical compression of the cingulate gyrus, by inducing vascular lesions via
trauma to the adjacent pericallosal artery or by direct contusion or even a slicing of the corpus callosum as it collides against the posterior ex-
Fig. 2.31 - Shearing injury. The FLAIR MRI sequence (b) is more sensitive than the T2-weighted FSE sequence (a) in iden- tifying even very small junction lesions (reprinted by M. Gal- lucci et al., 9)
a
b
tent of the free edge of the callosum. The neu- ropathological picture is always that of a tran- section of the involved nerve fibres. The most
common sites involved in this pathological process are the splenium and posterior body of the corpus callosum, in large part because of the proximity of these areas to the posterior free margin of the falx cerebri. The MR semei- otics are similar to those described for junc- tional lesions (Fig. 2.30).
Brainstem lesions constitute the clinically most serious form of axonal damage and can be divided into primarily non-haemorrhagic neu- ronal and primarily haemorrhagic microvascu- lar lesions. Neuronal lesions, which are usually more severe from a clinical point of view, most frequently represent a transection of nerve fi- bres of the ascending sensory pathways, or more rarely, injuries of the fibres of the descending motor pathways. They are generally caused fol- lowing torsion trauma due to the same mecha- nisms described for commissural lesions. These lesions tend to be round morphologically. The MR signal semeiology is typically identical to that seen in junctional lesions. These lesions are usually located in the dorsolateral portion of the pons and midbrain (Fig. 2.30). Primary haem- orrhagic microvascular brainstem trauma, or Duret lesions, occur following the tearing of the perforating branches of the basilar artery or the proximal portion of the posterior cerebral ar- tery. The most frequent location for this type of haemorrhagic lesion is the ventral midline por- tion of the brainstem. This type of lesion is of- ten associated clinically with a “locked-in” syn- drome. With regard to the MR signal character- istics, the same comments apply as those made above for haemorrhagic lesions.
The superior sensitivity of MR as compared to CT in the identification of brainstem lesions is further increased because of the obscuring effects of the beam hardening artefacts present on CT within the posterior fossa (Fig. 2.33).
Primary extraaxial traumatic lesions
Primary extraaxial traumatic lesions account for approximately 45% of all head injuries. In- cluded in this category are subarachnoid haem- orrhage, epi- and subdural haematomas, and subdural hygromas.
Fig. 2.32 - Haemorrhagic shearing injuries. Unlike FSE T2- weighted sequences, gradient recalled echo sequences which are more sensitive to T2* effects, are able to distinguish the presence of a haemorrhagic components within very small grey- white matter junction shearing injuries. [a) T2-weighted MRI;
b) T2*-weighted MRI].
Subarachnoid haemorrhage consists of local- ized or diffuse haemorrhage within the sub- arachnoid space. These haemorrhages rarely oc- cur as isolated lesions but instead are more fre- quently associated with extraaxial haematomas or cerebral laceration-contusions (3, 4, 28). A distinction can be made between primary and secondary subarachnoid haemorrhage. Primary subarachnoid haemorrhage results from the subarachnoid rupture of small pial-cortical sur- face vessels, from the tearing of bridging veins traversing the subarachnoid space or from the posttraumatic rupture of arteriovenous malfor- mations or cerebral aneurysms. Secondary sub- arachnoid haemorrhage, on the other hand, is caused by the rupture of a subdural haematoma or a cerebral laceration-contusion into the sub- arachnoid space.
Routinely, CT is undoubtedly the examina- tion of choice when subarachnoid haemorrhage is suspected. However, MR is more sensitive than CT in detecting the smaller subarachnoid bleeds, because it is able to identify the pres- ence of small quantities of blood in the sub- arachnoid space if FLAIR sequences are used.
In cases of massive primary subarachnoid haemorrhage, MR examinations performed in the acute phase combined with MR angiogra-
phy (MRA) can potentially identify the AVM or aneurysm responsible for the bleed.
Subdural haematomas can be either primary, due to the rupture of superficial bridge veins or small, superficial cortical arteries, or sec- ondary, when the presence of an intracerebral haematoma with a leptomeningeal margin re- sults in a rupture of the blood into the subdur- al space (8). The rapid diagnosis of acute sub- dural haematomas in head injury patients is quite important, as surgical evacuation within the first 4 hours reduces the death rate in such patients by 30%. Yet again, when acute sub- dural haematoma is suspected, CT represents the technique of choice in head trauma pa- tients. Subdural haematomas that are visual- ized on MRI but not seen on CT are usually small and of no surgical significance (Fig.
2.34). However, the discovery of small subdur- al haematomas can be important in distin- guishing these lesions from other types of pathology, or alternatively in cases of legal or forensic interest, in which case MRI is the im- aging technique of choice because of its higher relative sensitivity overall.
Acute subdural haematomas reveal charac- teristic MR signal, including isodensity on T1- weighted sequences and hyperintensity on PD and T2-weighted sequences indicating princi- pally oxyhaemoglobin. The presence of clots within the subdural haematoma can be respon- sible for an inhomogeneity in the signal, these clots appearing hyperintense on T1 sequences and hypointense on T2-dependent sequences (i.e., the combined effects of clot consolidation and deoxyhaemoglobin).
Extradural or epidural haematomas are bi- convex haemorrhagic collections with smooth, distinct margins, situated between the internal table of the skull and the dura mater. These haematomas are most commonly caused by the rupture of the middle meningeal artery or one of its branches, or less frequently, rupture of the posterior meningeal artery, one of the dural venous sinuses or the diploic vessels of the cal- varia. Epidural haematomas are usually en- countered at temporobasal or temporoparietal sites, and in most cases (85-90%) are associat- ed with fractures of the adjacent skull. With re-
Fig. 2.33 - Shearing injury. Identified on this T2-weighted FSE sequence is a small posttraumatic lesion of the posterior medul- la oblongata, following attempted suicide by hanging.
Fig. 2.34 - Haemorrhagic contusion, subdural haematoma and skull fracture. The CT examination (a, b) shows a haemorrhagic con- tusion associated with a small subdural haematoma (small solid arrows). The MRI conducted using T1- (c) and T2- (d) weighted se- quences more clearly shows the thin subdural haematoma located adjacent to the overlying bone (small solid arrows). CT (b) allows more direct identification of small associated bony fractures, although fractures may also be visible on MRI (arrowheads). (reprinted from M. Gallucci, et al., 9) [see Fig. 2.36].
a
b
c
d
gard to the identification of extradural haematomas, the sensitivity MRI is similar to that of CT, but MRI is more accurate. This is principally due to the ability of obtaining mul- tiplanar acquisitions on MRI that inherently have a greater probability of visualizing even
small bleeds without artefact, and the difficulty encountered with CT in the detection of thin hyperdense collections of blood adjacent to bone (Fig. 2.29). However, CT does permit the simple identification of associated bone frac- tures when present. In any case, the thin ex-
Fig. 2.35 - Left-sided cerebellar contusion. An MRI examination conducted in the hyperacute phase shows the presence of oedema in a left-sided cerebellar contusion (a, b). The MRI follow up conducted 15 days later documents the partial resolution of the oede- ma within the cerebellar contusion on the left. (a, b) (reprinted from M. Gallucci, et al., 9)
a
b
c
d
tradural collections that may be missed on CT but are visible on MRI do not require surgical intervention. Once again, legal and forensic in- terests may demand an MR examination in spite of these observations.
A tear of the leptomeninges and inner or meningeal layer of the cranial dural mater dur- ing head trauma can result in the passage of CSF into the subdural space. The resulting for- mation of a fluid collection, that is isointense with the fluid in all MR acquisitions, is known as a subdural hygroma (8). Subdural hygromas typically are not large and therefore produce little mass effect upon the underlying cerebral tissue, although exceptions are occasionally en- countered.
Acute secondary traumatic lesions
These secondary complications arise at a fairly early stage in the evolution of head trau- ma, assume their own clinical importance and can eventually come to dominate the patient’s clinical condition.
Cerebral swelling of a diffuse nature occurs in association with 10-20% of all severe head injuries, having a greater incidence in children.
This type of brain swelling usually occurs in connection with a primary traumatic lesion of the cranium that in 85% of cases is a subdural haematoma. Diffuse cerebral swelling occurs as an isolated lesion in only 5% of cases of head trauma. It typically appears within 4-48 hours of the traumatic incident and represents an in- crease in cerebral volume in part due to diffuse vascular paralysis (i.e., loss of cerebrovascular self-regulation) with a consequential increase in
Fig. 2.36 - Posttraumatic right-sided cerebellar infarct. T2- weighted MRI shows the presence of a cerebellar infarct in the territory of the right posterior inferior artery. The frontal MRA examination reveals a thinned, string-like incomplete MRI flow signal within the right vertebral artery. A T1-weighted scan of the neck, carried out 4 days after the traumatic event, shows a dissection of the right vertebral artery with intramural hyperin- tensity (open arrows) due to the presence of methaemoglobin.
(reprinted from M. Gallucci, et al., 9)
the cerebral blood pool. Cerebral swelling is therefore different from cellular or cytotoxic oedema (see below) of the cerebrum, which is caused by an increase in free water in the intra- cellular compartment, or vasogenic edema which is a result of leakage of fluid from the
vascular system into the extracellular compart- ment (21, 7, 27). The swelling can be mono- hemispheric or bihemispheric. Monohemi- spheric swelling tends to a more serious prog- nosis due to the greater incidence of internal cerebral herniation associated with this sub- type. The overall mortality in cases of diffuse cerebral edema is approximately 50%. CT and MRI have similar sensitivity in diagnosing cere- bral swelling. From a semeiological point of view, both techniques show a compression of the lateral ventricles and an effacement of the cerebral sulci over the cranial convexity. In the absence of complications or associated trau- matic effects, no signal alteration on MRI can be identified in these cases on T2-weighted se- quences.
Whether focal or diffuse, as it is linked to an increase in the amount of free water, cytotoxic and vasogenic cerebral oedema translates on MRI into a reduction of signal intensity on T1- weighted images and in an increase in signal in- tensity on PD- and T2-weighted sequences.
The gradual resolution of the oedema in the days following the trauma can be well docu- mented on MRI, with a parallel resolution of the T2-dependent hyperintensity of the oede- matous brain tissue.
Vascular lesions. Trauma can cause the lacer- ation, compression or dissection of the large vessels in the head and neck or of the intracra- nial vessels, and consequentially this can result in ischaemic or haemorrhagic lesions of the brain (9, 11, 13). The semeiotics of these sec- ondary ischaemic traumatic lesions is no differ- ent than that described for primary ischaemic lesions. In such cases, MRI has an advantage over CT in that it is able to integrate into one basic examination the documentation of both the ischaemic lesion by means of MRI, as well as the responsible abnormality of the vascular system by means of MRA (Fig. 2.36).
Internal cerebral herniations typically occur in association with cases of posttraumatic haemorrhage in which the collections exert sufficient mass effect to result in the dramatic dislocation of brain tissue within the skull.
Herniations are usually transfalcian, transten- torial, transphenoid or transalar in location,
Fig. 2.37 (cont. of 2.36) - Basal ganglia cytotoxic oedema. T2- weighted MRI shows hyperintensity within the basal ganglia and cerebral cortex signal consistent with cytotoxic oedema in a case of brain death. (reprinted from M. Gallucci, et al., 9).
however, internal herniations can and do occur in the cerebellar tonsils and inferior vermis downwardly through the foramen magnum.
The multiplanar capability of MRI and its more clear visualization of the posterior fossa structures routinely make it more informative than CT for the evaluation of this type of pathology.
Brain death. The literature contains some experiences with the MR angiography tech- nique that demonstrate the absence of flow in the intracranial circle in the event of brain death (9, 21).
Our experience, associated with a review of the literature, identifies a number of imaging parameters seen in cases of brain death follow- ing cranial trauma. These findings include a loss of the gray-white matter distinction, low signal intensity within the cerebral cortex and basal ganglia related to brain oedema (Fig.
2.37), absence of vascular enhancement after IV contrast medium administration attributed to the absence of intracerebral flow, and an ab- sence of MR flow signal in the main intracranial arterial structures (e.g., carotid siphons, middle cerebral arteries).
CONCLUSIONS
The observations in this chapter show that MRI is generally more sensitive than is CT in identifying the acute primary and secondary in- juries associated with cranial trauma. However, this is counterbalanced by the intrinsic limits of the technique and the difficulties inherent in the clinical management of the patient with acute head trauma.
These limitations have recently been at least partially overcome by technological progress with regard to reduced MR scanning times and the better differentiation of haemorrhagic and non-haemorrhagic lesions on MRI in the acute phase of head trauma. This is in turn counterbal- anced by the insurmountable incompatibility of many metallic life support appliances with the magnetic fields present in and near the MRI unit.
In summary, MRI is now more widely ap- propriately utilized in acute cases of cranial
trauma than it was even a few years ago. Nev- ertheless, CT still remains the undisputed tech- nique of choice for use in most acute phase trauma patients due to the relative rapidity with which it can be performed, the ready availabil- ity of the equipment required and its compati- bility with the life support systems often re- quired by trauma patients.
REFERENCES
1. Atlas SW, Mark AS, Grossman RI et al: Intracranial he- morrhage gradient-echo MR imaging at 1.5 T. Comparison with spin-echo imaging and clinical applications. Radio- logy 168:803-807, 1988.
2. Cirillo S, Simonetti L, Di Salle F et al: La RM in ematolo- gia con particolare riguardo all’emorragia intracranica. In:
Trattato delle malattie del sangue. Ed. P. Larizza pp. 563- 568. Piccin, Padova, 1990.
3. Cirillo S, Simonetti L, Di Salle F et al: La RM nell’emorra- gia sub-aracnoidea: Parte 2, studio in vivo. Rivista di Neu- roradiologia, 2:219-225, 1989.
4. Cirillo S, Simonetti L, Elefante R et al: La RM nell’emor- ragia sub-aracnoidea: Parte 1, studio in vitro. Rivista di Neuroradiologia, 2:211-217, 1989.
5. Elefante R, Cirillo S, Simonetti L et al: Confronto TC ed RM in patologia encefalica. In: La TC in Neuroradiologia, Microart’s, Genova pp. 169-174, 1991.
6. Espagno J, Manelfe C, Bonsique JY et al: Interet et valeur prognostique de la TDM en traumatologie cranio-cerebra- le. J Neuroradiol 7:121-32, 1980.
7. Espagno J, Tremoulet M, Espagno C: The prognosis of brainstem lesions in patients with recent head injuries. J Neurosurg Sci 20:33-50, 1976.
8. Fohlen ES, Grossman RI, Atlas SW et al: MR characteri- stics of subdural hematomas and hygromas at 1.5 T. Am J Neuroradiol 10:687-693, 1989.
9. Gallucci M, Bozzao A, Maurizi Enrici R: La Risonanza Magnetica nei traumi acuti. In: Pellicanò G, Bartolozzi A (eds): La TC nei traumi cranio-encefalici. Ed. del Centau- ro, Udine, pp. 129-153, 1996.
10. Gentry LR, Godersky JC, Thompson B et al: Prospective comparative study of intermediate-field MR and CT in the evaluation of closed head trauma. Am J Neuroradiol 9:91- 100, 1985.
11. Gentry LR: Head trauma. In: Atlas S.W. Magnetic Reso- nance imaging fo the brain and spine. Lippincott Raven Press, Philadelphia 661-647, 1996.
12. Gentry LR: Imaging of closed head injury. Radiology 191:1-17, 1994.
13. Gentry LR, Godersky JC, Thompson B: MR imaging of head trauma: review of distribution and radiopathologic feature of traumatic lesions. Am J. Neuroradiol 9:101-110, 1985.
14. Gallucci M, Splendiani A, Bozzao A: Valutazione compa- rata RM-TC nel traumatizzato cranico. Eido Electa, 11-21, 1985.
15. Hesselink JR, Dowel CF, Healy ME et al: MR imaging of brain contusions: a comparative study with CT. Am J Neu- roradiol 9:269-278, 1985.
16. Kelly AB, Zimmerman RD, Snow RB et al: Head trauma:
comparison of MR and CT-experience in 100 patients.
AJNR 9:269-78, 1988.
17. Lanksch W, Grumne T, Kazner E: Correlation between cli- nical symptoms and CT findings in closed head injuries. In:
Frowein, Wilcke, Karini (eds): Advances in neurosurgery Vol 5: Head injuries. Springer-Verlag, Berlin, pp. 27-29, 1975.
18. Levin HS, Amparo E, Eisemberg HM et al: MR and CT in relation to the neurobehavioral sequelae of mild and mo- derate head injuries. J Neurosurg 6:706-713, 1987.
19. Ogawa T, Sekino H: Comparative study of MR and CT scan in cases of severe head injury. Acta Neurochir 55:8- 10, 1992.
20. Orlandi B, Cardone G, Minio Paluello GB et al: Sequenze standard e sequenze veloci: bilancio di utilità su fantoccio ed in vivo. In: Neuroradiologia 1995, a cura di L. Bozzao.
Ed del Centauro, Udine, 161-166, 1995.
21. Osborn AG: Diagnostic Neuroradiology. Mosby Year Book, St. Louis 1994. Cap. 8.
22. Passariello R: Elementi di tecnologia in radiologia e dia- gnostica per immagini. Cromac, Roma 1990.
23. Pierallini A, Pantano P, Zylberman R et al: Valutazione RM con sequenze FLAIR e FFE del carico lesionale in pazien-
ti affetti da gravi traumi cranici chiusi. Rivista di Neurora- diologia 10:57-60, 1997.
24. Piovan E, Beltramello A, Alessandrini F et al: La Risonan- za Magnetica nei traumi sub-acuti e cronici. In: Pellicanò G, Bartolozzi A (eds): La TC nei traumi cranio-encefalici.
Ed. del Centauro, 153-171, Udine 1996.
25. Scotti G, Pieralli S, Righi C et al: Manuale di neuroradio- logia diagnostica e terapeutica.
26. Sklar EM, Quencer RM: MR application in cerebral injury.
RCNA 30(2):353-56, 1992.
27. Splendiani A, Gallucci M, Bozzao A et al: Ruolo della Ri- sonanza Magnetica nello studio dei traumi cranici acuti e cronici: studio comparativo con tomografia computerizza- ta. In: Pardatscher K.: Neuroradiologia 1989. Ed. del Cen- tauro, Udine 1989.
28. Yoon HC, Lufkin RS, Vinuela F et al: MR of acute suba- rachnoid hemorrhage. Am J Neuroradiol 9:404-5, 1988.
29. Zimmerman RA, Bilaniuk L, Gennarelli T et al: Cranial CT in the diagnosis and management of acute head trau- ma. Am J Radiol 131:27-34, 1975.
30. Zimmerman RD, Heier LA, Snow RB et al: Acute intra- cranial hemorrage. Intensity changes on sequential MR scans at 0.5 T. Am J Neuroradiol 9:47-57, 1985.
